Engineering that ships.
Sixteen years of building production software across AI, energy, fintech, and IoT. I take complex technical problems and turn them into working systems.
AI and machine-learning oriented development
I’ve been building AI systems professionally since well before the ChatGPT era. My work in this space goes back to corrosion detection using computer vision in 2016, and I’ve shipped production AI for energy monitoring, fintech advisory, and ESG compliance since then.
Today I work primarily with LLM-based architectures — agentic workflows built on LangGraph and LangChain, with models from Claude and Gemini. I also have deep experience with classical ML: PyTorch, TensorFlow, and OpenCV for computer vision tasks.
The references below document my AI work years before generative AI became mainstream. That matters because it shows this isn’t a bandwagon — it’s a domain I’ve been investing in for a long time.
Power and utilities — from tender to exit
This was the biggest chapter from the Broentech Solutions days. We won a tender from Glitre Energi to build an AI-native web application for monitoring hydropower infrastructure. That contract turned into a product, the product attracted attention, and in 2024 it led to an acquisition by Captiva.
The system handled real-time sensor data from power stations, used machine learning for anomaly detection, and gave operators a unified dashboard for their entire portfolio. It was a full-stack effort: React frontend, Python backend, Kubernetes infrastructure, PostgreSQL with TimescaleDB for time-series data.
This track record matters because it’s a complete arc — winning a public tender, building the product, scaling it, and completing a successful exit. It shows I can take a project from zero to acquisition.
Startups and technical leadership
I’ve founded two companies: Broentech Solutions AS (2012–2024, acquired) and Broentech Sentinel AS (2026–present). Eleven years as founder and CEO taught me that building a company is a fundamentally different skill from building software — and you need both.
I’m also an active investor and technical co-founder. At Investwiser, I serve as CTO and developer, building the AI-powered investment advisory platform. The first version used a self-hosted LLM running on GPU nodes in Kubernetes with Nvidia CUDA drivers and a model from Hugging Face — real inference infrastructure, not just API calls.
Not every venture succeeds. Vedbjørn, a firewood logistics platform, didn’t make it. But the experience of building it — the technical architecture, the go-to-market challenges, the decision to shut down — is part of what I bring to every new project. I also built AINA, an ESG compliance platform, which is my most recent product venture.
Advisory and lecturing
I’ve been involved in multiple technical due-diligence cases, advising both startups and enterprises on technology strategy, architecture decisions, and build-vs-buy trade-offs.
Since 2018, I’m a recurring guest lecturer on AI at the University of South-Eastern Norway (USN). Teaching forces you to explain things clearly — a skill that transfers directly to client work.
Computer vision
My computer vision work spans C++, Qt, OpenCV, and Python. I’ve built image processing pipelines for industrial inspection and published open-source tools for spatial filtering.
C++ and Qt — Nokia certified
Before the web ate everything, I cut my teeth on native desktop and embedded applications with C++ and Qt. Around 2010 I earned both the Nokia Qt Developer and Nokia Qt Specialist certifications — back when Nokia ran the Qt certification programme. It’s old, but it’s real: formal validation of deep knowledge in cross-platform C++ development, event-driven architectures, and the Qt framework internals.
That background still pays off. Qt shows up in industrial software, embedded systems, and anywhere performance matters more than convenience. I’ve built open-source tools with it — including sauto, a library for sensor-triggered automation — and the C++ discipline carries into everything I write today.
Technologies I work with
DevOps
Languages
Frameworks
AI / ML
Databases
Companies I've worked with
Over sixteen years, I've delivered projects for companies across maritime, energy, healthcare, and media.
Kongsberg Maritime AS
GE Healthcare — Vingmed
Kystverket
Glitre Energi AS
(now Å Energi)
Hydro Energi
Schibsted
(via Duplo Media)
Novelda
Ablecon AS
Have a technical challenge?
I’m available for consulting, development, and advisory work. Let’s talk about what you’re building.